Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Recent regression in spurious invalidations. #51146

Closed
KristofferC opened this issue Sep 1, 2023 · 0 comments · Fixed by #51153
Closed

Recent regression in spurious invalidations. #51146

KristofferC opened this issue Sep 1, 2023 · 0 comments · Fixed by #51153
Assignees
@vtjnash
Labels
compiler:latency Compiler latency kind:regression Regression in behavior compared to a previous version
Milestone
1.10

Comments

@KristofferC
Copy link
Sponsor Member

KristofferC commented Sep 1, 2023 •
edited
Loading

Running with Pkg out of the sysimage (using this commit 412c8f4) and the following Pkg commit JuliaLang/Pkg.jl@f2d8b64 there is a recent regression where the precompilation workload in Pkg seems to get heavily invalidated and becomes basically worthless:

# Good
julia> @time Pkg.add("Example")
...
  1.735661 seconds (3.62 M allocations: 237.936 MiB, 7.02% gc time, 45.09% compilation time)

# Bad
julia> @time Pkg.add("Example")
...
 11.234634 seconds (4.29 M allocations: 307.999 MiB, 91.73% compilation time: 73% of which was recompilation)

This recompilation seems spurious since there are no (AFAIK) other packages getting loaded that would cause this.

This was bisected to #51036 cc @vtjnash

Marking for 1.10 since the offending commit is marked for backporting.
@KristofferC KristofferC added this to the 1.10 milestone Sep 1, 2023
@brenhinkeller brenhinkeller added compiler:latency Compiler latency kind:regression Regression in behavior compared to a previous version labels Sep 1, 2023
@vtjnash vtjnash mentioned this issue Sep 1, 2023
Merged
N5N3 pushed a commit that referenced this issue Sep 3, 2023
@vtjnash @DilumAluthge
fix debug typo in "add missing invoke edge for nospecialize targets (#…
da86735 
…51036)" (#51153)

Causes `matches` to get replaced with `MethodMatch` instead, which then
later will fail to match with the expected value.

Fixes #51146

Co-authored-by: Dilum Aluthge <dilum@aluthge.com>
KristofferC pushed a commit that referenced this issue Sep 15, 2023
@vtjnash
fix debug typo in "add missing invoke edge for nospecialize targets (#…
0e7eed8 
…51036)" (#51153)

Causes `matches` to get replaced with `MethodMatch` instead, which then
later will fail to match with the expected value.

Fixes #51146

Co-authored-by: Dilum Aluthge <dilum@aluthge.com>
(cherry picked from commit da86735)
nalimilan pushed a commit that referenced this issue Nov 5, 2023
@vtjnash @nalimilan
fix debug typo in "add missing invoke edge for nospecialize targets (#…
e509c33 
…51036)" (#51153)

Causes `matches` to get replaced with `MethodMatch` instead, which then
later will fail to match with the expected value.

Fixes #51146

Co-authored-by: Dilum Aluthge <dilum@aluthge.com>
(cherry picked from commit da86735)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@vtjnash vtjnash

Labels
compiler:latency Compiler latency kind:regression Regression in behavior compared to a previous version
Projects
None yet
Milestone
1.10
Development

Successfully merging a pull request may close this issue.

fix debug typo in "add missing invoke edge for nospecialize targets (#51036)" JuliaLang/julia
3 participants
@vtjnash @KristofferC @brenhinkeller